Does Google Drive / Dropbox do a FULL backup of ERPnext?

If i backup with the integrated Google Drive or DropBox function, will this be a full backup of everything? All data and settings?

ERPNext takes an entire system backup and stores it as a SQL file. The integrations simply push the generated backup to a service. Check out the documentation for more details.

1 Like

In my experience, the automated system backup is ONLY a backup of the database (I am using v10 and v11). While the database has all of the critical information to restore a system, is is by no means complete.

There are associatef files that are stored in other public and private directories that need to be backed up as well. One example would be the pictures of items in your Items data are not stored in the database.


1 Like

There are different kinds of data:

  1. Your site’s database
  2. Your site’s public files (for example, images on your website)
  3. Your site’s private files (for example, purchase invoice PDFs)
  4. Server configuration (frappe-bench/config/*)
  5. Installed apps and their state (which commit is checked out)
  6. State of your server (installed dependencies, etc)

(1) is typically included in external backups, (2) and (3) are optional. You have to take care of (4), (5) and (6) yourself, if you do self-hosting.

For example, if you have a database backup of a v9 instance, you will have great difficulty restoring it now, because the dependencies changed and it’s close to impossible to set up a new working server with all the old software required (here’s how). However, if you have a full server backup, there shouldn’t be many problems.


Something to consider…

1 Like

site_config is part of backup now

1 Like

Both Google Drive and Dropbox backups take a file backup as well. Not sure about the S3 one.

S3 can be mounted as a drive as well as any Cron job could upload it using the AWS CLI

I agree with what @rmeyer wrote. The built-in backup routines do not cover everything.

If you want a FULL backup (as the original poster mentioned), I suggest writing your own shell scripts.

And even then. You should consider things like OS dependencies (MariaDB version, Node version). Just because you have a pristine backup, doesn’t mean it would actually function on <some_machine_somewhere>

The only good backup plan is one you’ve tested. Simulate a disaster, and then make sure you can successfully restore.